In Experiment One, mock jurors viewed a videotaped recreation of a court trial in which the age of the prosecution's key witness was presented as 8, 21, or 74 years of age. Contrary to prior research in this area, the testimony of the 8-year-old witness was rated as more credible than identical testimony given by the 21-year-old witness. The elderly witness was also viewed as more credible than the young adult witness, but less so than the child witness. These findings were replicated in Experiment Two, where mock jurors read a written transcript of the same trial that was presented via videotape in Experiment One. In Experiment Three, a survey was taken of mock jurors' beliefs about age differences in eyewitness ability. In general, mock jurors were found to hold a negative stereotype of the child witness. These findings are discussed in terms of current theory and research on juror reactions to the child witness, and the more general issue of how stereotypes influence impression formation and social judgement. 3 tables, 32 references. (Author abstract)